Phonological memory is the capacity to remember speech sound sequences that are brand new.

J Exp Child Psychol. Author manuscript; available in PMC 2014 Jun 27.

Published in final edited form as:

PMCID: PMC4073230

NIHMSID: NIHMS230461

Abstract

The relation of phonological memory to language experience and development was investigated in 41 Spanish-English bilingual first language learners. The children’s relative exposure to English and Spanish and phonological memory for English-like and Spanish-like nonwords were assessed at 22 months; their productive vocabulary and grammar in both languages were assessed at 25 months. Phonological memory for English- and Spanish-like nonwords were highly correlated, and each was related to vocabulary and grammar in both languages, suggesting a language-general component to phonological memory skill. In addition, there was evidence of language-specific benefits of language exposure to phonological memory skill and of language-specific benefits of phonological memory skill to language development.

Keywords: Phonological memory, Phonological representations, Vocabulary development, Grammatical development, Bilingual development, Language specificity, Language input

A substantial body of evidence from the study of first and second language acquisition argues that phonological memory (i.e., the capacity to remember sequences of sounds) is a component of the human language acquisition capacity. Children and adults who have better phonological memory skills acquire language more rapidly than children and adults who are less able to remember novel auditory stimuli (Gathercole, 2006). Phonological memory skills appear to have both a general auditory memory component, which operates over all speech-like stimuli without drawing on information in long-term memory, and a component that makes use of knowledge based on prior language experience (Vallar, 2006). Evidence of the influence of language experience includes findings that children show better memory for sound sequences in real words than nonwords (Chiat & Roy, 2007, Snowling, 1981), better memory for high frequency than low frequency sound sequences in nonwords (Edwards, Beckman, & Munson, 2004; Munson, Kurtz, & Windsor, 2005), and better memory for sound sequences that conform to the phonology of their own language than for sequences drawn from a foreign language (Thorn & Gathercole, 1999).

The effect of language experience on phonological memory has implications for the process of phonological memory development and its role in bilingual development. The phonological memory skills of children exposed to two languages might include two different language-specific components, each drawing on knowledge of one of their languages. These two knowledge bases might develop at different rates if the children’s exposure to one language is greater than their exposure to the other. Furthermore, to the degree that the value of phonological memory to subsequent language development depends on a language-specific capacity to store sound sequences, bilingually-developing children’s phonological memory skills in each language should have language-specific benefits. The present study was designed to test these hypotheses. In the following sections we first review the literature that establishes the relation of phonological memory skill to vocabulary and grammar in monolingual development and in second language acquisition. We then outline the theoretical issues and evidence regarding the influence of language exposure and familiarity on phonological memory skill. Last, we review the literature on bilingual development that is relevant to the hypothesis that bilingual children acquire language-specific phonological memory skills as a result of language exposure and that the value of those skills to subsequent language development is, in part, language specific.

Phonological Memory and Vocabulary

Phonological short term memory has been demonstrated to be related to vocabulary knowledge and vocabulary development in first language learning, even after considering the effects of age and nonverbal intelligence (Gathercole, Hitch, Service, & Martin 1997; Gathercole, Willis, Emslie, & Baddeley, 1992). Relations between phonological memory skill, measured as accuracy of nonword repetition, and vocabulary have been demonstrated in samples of children between 20 months and 8 years (Chiat & Roy, 2007; Gathercole & Adams, 1993; Gathercole et al., 1992; Hoff, Core, & Bridges, 2008).

Phonological memory is important to vocabulary learning in second language learning as well. Among adolescent foreign language learners, repetition accuracy for English-like stimuli has been found to be related to success at English vocabulary learning (Service & Kohonen, 1995). As with first language acquisition, phonological memory seems to play a role at the early stages of second language learning. For example, the relation between English nonword repetition and speed of learning English vocabulary was significant for Cantonese seventh graders (M age = 12 years) learning English at school who had low English vocabulary skills but not for those who had high English vocabulary skills (Cheung, 1996).

Phonological Memory and Grammar

Phonological short term memory has been also associated with grammatical development in first and second language acquisition. Adams and Gathercole (1995) found that 3-year-olds with good nonword repetition skills differed from children poor at nonword repetition with respect to the variety of vocabulary, the length of the utterances, and the complexity of the syntax used in spontaneous speech. Adams and Gathercole (1996) also found that nonword repetition skills were associated with the length of sentences and the amount of detail in narrations produced by 4-year-old children, after controlling for vocabulary knowledge, age, and nonverbal intelligence. Finally, Adams and Gathercole (2000) found that 4-year-old children with good nonword repetition skills produced longer utterances and more varied syntactic constructions than did children of the same age and similar nonverbal abilities who had poor nonword repetition skills.

In the realm of second language learning, French and O’Brien (2008) examined the relation between phonological memory and grammar learning in a group of 11-year-old French children enrolled in a 5-month intensive English program. Phonological memory was measured at the beginning (i.e., Time 1) and at the end of the program (i.e., Time 2) with English nonwords and with Arabic words, which were functionally nonwords to the children. Both English nonword and Arabic word repetition tasks measured at Time 1 explained a significant amount of variance in grammar at Time 2 above and beyond the variance explained by vocabulary knowledge (at Times 1 and 2), non-verbal intelligence, and grammatical knowledge at Time 1. Furthermore, although English nonword and Arabic word repetition accuracy were highly correlated within and across measurement times, the former increased over time while the latter did not. This was attributed to the lack of influence of lexical knowledge on Arabic words relative to English nonwords. Thus, repetition accuracy for Arabic words (in nonArabic speakers) was argued to be a more sensitive measure of phonological memory.

The Effect of Language Familiarity on Phonological Memory

Although phonological memory was originally hypothesized to be an unlearned cognitive capacity to process and store phonological information (Baddeley, Gathercole, & Papagno, 1998), phonological memory has more recently been argued to be affected by extrinsic factors such as language input (MacDonald & Christiansen, 2002) and to reflect the quality of children’s phonological representations (Gathercole, 2006). In support of this experience-dependent view of phonological memory, Snowling (1981) showed that dyslexic readers (M age= 12 years) and reading-level matched normal readers (M age= 8.4 years) were better at repeating real words than nonwords, and Gathercole, Willis, Emslie and Baddeley (1991) found wordlikeness effects on nonword repetition accuracy in children from 4 to 6 years. Several studies have found that children are better at repeating nonwords containing high frequency phonemes than low frequency phonemes (Coady & Aslin, 2004; Edwards et al., 2004; Munson et al., 2005), and Messer, Leseman, Boom and Mayo (2010) found in young second language learners that this benefit of high phonotactic probability to nonword repetition was greater in children who were more familiar with the language. Adults have shown better memory for sound sequences that conform to the language they know than for sound sequences in a foreign language (Service & Kohonen, 1995; Thorn & Gathercole, 1999). In domains other than language, there is evidence that knowledge of a particular domain results in the development of a representational system that improves memory performance in that domain (Schneider, Bjorklund, & Maier-Brückner, 1996), and phonological memory seems to similarly benefit from the support of a knowledge base. If phonological memory is supported by experience-dependent language knowledge, then bilingually-developing children, who have different amounts of experience in knowledge of each of their languages, might then have different levels of phonological memory skill in each language. These different levels of phonological memory skill might then, in turn, provide different levels of support for the development of each language.

Phonological Differentiation in Bilingual Development

Although young bilingual children’s vocabulary knowledge has been demonstrated to differ between languages as a function of input (Pearson, Fernández, Lewedeg, & Oller, 1997), no studies have directly addressed the question of whether bilingually-developing children show different levels of phonological knowledge in their two languages as a result of differences in input. There is evidence that bilingual input affects early phonological development, and there is evidence that bilingually-developing children can have separate, if not completely autonomous, phonological systems. For example, infants as young as 10 or 12 months who are exposed to two languages retain the ability to hear phonemic contrasts in both languages, which monolingually-exposed infants lose (Bosch & Sebastian-Gallés, 2003; Burns, Yoshida, Hill, & Werker, 2007). French-English bilingual 2-year-olds have distinct prosodic features in their French and English productions, although they also show crosslinguistic transfer from French into English in their repetition of English and French nonwords (Paradis, 2001). Also suggestive of separate systems is the finding from two Cantonese-English bilingual children, who were exposed to the second language at the age of 2 years, that they produced error patterns and phoneme simplifications that were specific to each language. In addition, their errors followed the phonotactic constraints of each of the languages, and shared phonemes did not always enter both productive systems at the same time (Holm & Dodd, 1999).

The present hypothesis does not require that bilinguals have two separate systems, only that they have different levels of familiarity with the sounds and/or sound sequences characteristic of each language. In order for that to be the case, of course, it is necessary that the two languages make use of different sounds and sound sequences. English and Spanish do make use of different phonemic inventories and have different phonotactic patterns (Goldstein, 2004). For example, Spanish words are more likely to end in vowels than consonants and never end in a consonant cluster, whereas word final consonants and consonant clusters are common in English. More broadly, differences between Spanish and English in word shapes, in word length, in stress patterns in multisyllabic words, and in the phonetic details of how individual consonants and vowels are articulated make Spanish and English sound very different. One source of evidence that these differences in the sounds of the two languages make encoding of speech difficult for speakers who are familiar with English but not Spanish is data from English monolingual and Spanish-English bilingual kindergarten children who were given the task of repeating fictitious Spanish names in English (e.g., how would you say “Parasco”—pronounced in Spanish—in English? (Oller, Cobo-Lewis, & Eilers, 1998). The monolingual children performed less well than the bilingual children, even though they did not have to produce any Spanish sounds; they only had to encode them as they were presented orally with Spanish phonology and map them onto their English equivalents.

The Present Study

The nature of the relations among language exposure, phonological memory and language development in early bilinguals has not been previously investigated. The present study tests the hypotheses that (1) children’s relative exposure to English and Spanish will be related to their phonological memory skills for English-like and Spanish-like stimuli and (2) that phonological memory skills will show language-specific relations to vocabulary and grammatical development.

Method

Participants

The participants were 41 Spanish-English bilingually-developing children (21 males, 20 females) who were born in the United States and were living in South Florida. According to parental report, they had been exposed to both English and Spanish since birth and the less frequently heard language constituted at least 10% of their input. The balance of language exposure ranged from 10% English and 90% Spanish to 90% English and 10% Spanish, with an average home language input of 49.9 % (SD = 29.1). All children were described by their parents as learning both languages and all children were producing words in both languages at 22 months.

Fifty percent of the mothers and 52% of the fathers were native Spanish speakers; 35.4% of the mothers and 37.5% of the fathers were native speakers of English; 12.5% of the mothers, and 8.33% of the fathers described themselves as native Spanish-English bilinguals; and 2.08% of mothers and 2.08% of the fathers of the participants were native speakers of a language other than English or Spanish. Eighty-nine percent of the native Spanish speakers were born in Spanish-speaking countries in South American and the Caribbean; 11 % were born in the U.S.

The children were 22 months at the start of the study (M = 22.78 months, SD = 0.39). All participants were full term at birth and had no history of medical or sensory problems. In addition, they had normal communicative and language development according to the criterion that they were above the recommended cutoff value of 35 for language delay in the communication section of the Ages and Stages Questionnaire (Squires, Potter, & Bricker, 1999) and were at least at the 10th percentile in at least one of their languages using the MacArthur-Bates Inventory Scales (Fenson et al., 1993; Jackson-Maldonado et al., 2003). On average, the children were at the 29th percentile for English vocabulary and the 42nd percentile for English grammar; they were at the 23rd percentile for Spanish vocabulary and the 49th percentile for Spanish grammar based on monolingual norms. These numbers are consistent with other evidence that bilingually-developing children show somewhat slower rates of development in each language—especially in vocabulary—than monolingual children (Bialystok, 2001; Bialystok & Feng, in press; Conboy & Thal, 2006; Marchman, Martínez-Sussmann, & Dale, 2004). Ninety-three percent of the children were white Hispanics, 2.4% were Hispanic of African descent, 2.4% were European American, and 2.4% belonged to other ethnicities. The socioeconomic status as assessed by the parents’ educational level was high: 89.5% of the mothers and 64.58% of the fathers had at least a 4-year college degree.

Design and Measures

The data were collected as part of a larger longitudinal study of early bilingual development. Children’s phonological memory skills and the balance of English and Spanish language exposure at home were measured at 22 months; productive vocabulary size and grammatical complexity of speech were measured at 25 months (M = 25.82 months, SD = 0.34).

Phonological memory

Nonword repetition tasks were used to measure phonological short term memory. The stimuli consisted of 12 English-like nonwords (kog, buice, jat, dook, challoon, pookie, kuppy, bicken, bajapop, tellina, lolemas, panaphone) and 12 Spanish-like nonwords (lan, trus, sen, pol, vato, meca, lesa, gache, gañeca, mullina, peballo, calota). These nonwords were constructed from real words taken from the MacArthur-Bates Communicative Development Inventory (Fenson et al., 1993) for children 16 to 30 months and its Spanish version, Inventario del Desarrollo de las Habilidades Comunicativas (Jackson-Maldonado et al., 2003). All the sounds that occur in the real words also occur in the nonword stimuli in the same word positions, and the nonwords followed the same phonotactic frames and stress patterns as the real words from which they were derived. Thus, the nonwords were phonologically like the words children acquire at this age in each language. We excluded late developing sounds, such as English /r/ and Spanish trilled /r/ as have others (e.g., Shriberg et al., 2009), because we wanted to ensure that children’s errors were based on repetition abilities rather than articulation abilities, to the degree possible. There were four one-syllable, four two-syllable, and four three-syllable nonwords in each language. The monosyllabic nonwords were constructed by interchanging the onsets and rhymes of the monosyllabic real words. The bisyllabic nonwords were formed by combining the onset of the first syllable of one real word with the rhyme and second syllable of another real word. The trisyllabic nonwords were a combination of the first, second and third syllables respectively of three different real words for English and the combination of the onset and nucleus of one real word with the second and third syllables of another real word for Spanish.

The procedure used to assess nonword repetition accuracy followed the procedure developed and validated by Hoff et al. (2008) to assess nonword repetition in children as young as 20 months. The nonwords were presented orally by an examiner who was a native speaker of the language that was the basis of the nonwords presented. The tasks using English-like and Spanish-like stimuli were administered on different days. For 34 children the sessions occurred in the children’s homes and for 7 they occurred in a laboratory playroom, depending on the caregivers’ preferences. Stimuli were presented in a standard way, embedded in a toy play activity. The examiner’s face was always visible to the child. The procedure was as follows: after a warm up period, children were presented with nonwords one at a time. The nonwords were accompanied by toys representing animals or people. Participants were told that the nonwords were the names of these toys, and were asked to repeat them back. For instance, children were told: “This guy is named Kog, can you say Kog?” The session started with two training trials. After two successful repetitions, the test stimuli were presented. Children were provided up to three presentations if they did not repeat the first or second presentation. The first repetition the child produced was scored. The children’s productions were recorded for later transcription. If a child failed to repeat six consecutive stimuli, the session was terminated. Only children who attempted to repeat at least 3 nonwords in either of the languages were included.

The accuracy of nonword repetition was measured by calculating the percentage of consonants presented that were repeated correctly (PCC), which is the most widely used measure of repetition accuracy (Coady & Evans, 2008). Percent correct was used rather than total number correct because of differences in the number of possible consonants presented that existed between the English-like and Spanish-like stimuli and because not all stimuli were presented to a child if he or she failed to repeat 6 consecutive stimuli. The scoring of repetition accuracy for English-like stimuli was done by an expert phonetician and a graduate student trained in phonetic transcription, both native English speakers. The scoring of repetition accuracy for Spanish-like stimuli was done by the expert phonetician, a fluent Spanish speaker, and two graduate students trained in phonetic transcription who were native speakers of Spanish. Disagreements were resolved by discussion until consensus was reached.

Language exposure

The percentage of the children’s language exposure that was in English and Spanish was estimated by the primary caregivers using the Home Language Environment Questionnaire (HLEQ). The HLEQ is 145-item interview protocol designed for this study, based on an instrument developed by Marchman et al. (2004). As part of the larger study from which these data are drawn, caregivers’ estimates of their children’s language exposure were validated against Language Diary data: 33 mothers kept language diaries in which they recorded the language their child was exposed to for each waking 30-minute period of 7 days. The correlation between caregivers’ estimates of the percent of their children’s language input that was in English and the number of hours per day of English exposure recorded in diary logs was high, r (29) = .71, p <.001. The diary data also revealed that English was more prevalent in children’s language exposure outside the home than inside, thus the estimates of home language exposure were underestimates of the proportion of the children’s language exposure that was in English (Place & Hoff, 2010). Including out of home exposure, the children in the present sample heard more English than Spanish, on average. The caregiver’s estimates of the relative amount of exposure do not, of course, capture all the variability in these children’s dual language experience. There was no measure of the absolute quantity of exposure to either language. However, the diary data suggest these are valid estimates of relative exposure and relative differences in exposure are likely to be related to differences in the quantity exposure, unless there are systematic differences in the amount of talk addressed to children related to the balance of English and Spanish addressed to children. We compared the levels of parental education among Spanish-dominant, balanced, and English-dominant households and found no difference. That finding argues against one potential confound between relative balance and absolute amount of child-directed speech.

Language development

Vocabulary and grammar in English and Spanish were assessed using caregiver report inventories, the MacArthur-Bates Communicative Development Inventory, CDI (Fenson et al., 1993), and the Spanish version, the Inventario del Desarrollo de las Habilidades Comunicativas-II, IDHC (Jackson-Maldonado et al., 2003). Vocabulary size was calculated from part I of the CDI and IDHC, Words Children Use. This part contains 680 words of different lexical categories including nouns, verbs, adjectives, prepositions, pronouns, and articles. Parents were asked to indicate the words they heard their children say.

The measure of grammatical development was calculated from part II of the CDI and IDHC, Sentences and Grammar. In this section, parents are presented with 37 pairs of sentences. Each pair contains one sentence that is more complex than the other. Parents are asked to select the sentence that is more representative of the sentences their children use. The number of times they chose the more complex sentence is calculated.

Procedure

Recruitment was carried out through electronic advertisements and advertisements in local magazines for parents of young children, through word of mouth, and by contacting parents at library events and preschools. In change for their participation children received a T-shirt and a toy, and caregivers a $25 gift card at each visit. The data for the present study were collected when the children were 22 months and 25 months. At 22 months, the nonword repetition tasks in English and Spanish were administered on separate days within a week of each other. The primary caregivers were contacted later when their children were 25 months to complete the communicative development inventories. When possible, English and Spanish native speakers completed the English and the Spanish versions, respectively; otherwise, a caregiver who was a proficient Spanish-English bilingual completed both inventories.

Results

The means and standard deviations for the measures of the relative amount of English language exposure at home, nonword repetition accuracy for English-like and Spanish-like stimuli at 22 months, and raw vocabulary and grammatical complexity scores in English and Spanish at the age of 25 months are presented in Table 1. Almost exactly half of the children’s language exposure at home was to English, on average, but recall, the Language Diary data available for a subset of these children indicated that language exposure outside the home is more dominated by English (Place & Hoff, 2010). Children’s scores for measures of phonological memory, vocabulary, and grammar were significantly higher in English than in Spanish, for nonword repetition accuracy, t (40) = 2.07, p < .05, for vocabulary size, t (40) = 2.98, p < .01, and for grammatical complexity of speech, t (40) = 2.22, p < .05, all two-tailed. This pattern is consistent with the children’s greater exposure to English than Spanish outside the home suggested by the diary data, but these findings should be interpreted cautiously because none of the instruments is calibrated to provide a basis for comparing the children’s skill across languages. (There were no gender-related differences on any of these measures.)

Table 1

Means (and Standard Deviations) for Bilingual Children’s English Language Exposure, Nonword Repetition Accuracy, Vocabulary Size, and Grammatical Complexity in English and Spanish (n=41)

The Relation of Language Exposure to Phonological Memory and Language Development

The percent of children’s home language exposure that was in English was positively related to their nonword repetition accuracy for English-like stimuli, r (n=41) = .26, p = .05, one-tailed, and unrelated to their nonword repetition accuracy for Spanish-like stimuli, r (n=41) = −.05, p = .37, one-tailed. Because repetition accuracy for English-like and Spanish-like stimuli were not independent (see Table 3), a clearer test of the hypothesis that language exposure benefits the development of language-specific phonological memory skills is the correlation between language exposure and repetition accuracy in one language, when the variance shared with repetition accuracy in the other language is held constant. They were both significant, r (n=41) = .50, p = .001, one tailed, for English nonword repetition and r (n=41) = −.45, p = .004, one-tailed, for Spanish nonword repetition. These findings indicate that with the shared variance in these two tasks removed, relative language exposure accounted for a significant 25% of the variance in performance on English-like stimuli and a significant 20% of the variance in performance on Spanish-like stimuli.

Table 3

Correlations among Nonword Repetition Accuracy at 22 Months and Vocabulary and Grammar at 25 Months, Within and Across Languages (n=41)

Correlations between relative language exposure and the measures of vocabulary and grammatical development assessed at 25 months are presented in Table 2. English exposure showed a strong positive relation to English vocabulary and grammar and a strong negative relation to Spanish vocabulary and grammar.

Table 2

Correlations between Percent of Language Exposure in English at 22 Months and Vocabulary Size, and Grammatical Complexity in English and Spanish (n=41)

The Relations of Phonological Memory to Language Development

Zero-order correlations among the phonological memory and oral language measures are presented in Table 3 (lines 1 – 6); these same correlations with effects of the relative amount of exposure to English held constant are presented in Table 3 (lines 7 - 12). In the zero-order correlations, phonological memory skill for English-like and Spanish-like stimuli were strongly and positively related (line 1, Table 3). Phonological memory skill for English-like stimuli was significantly related to English vocabulary and English grammar, but not to Spanish vocabulary or grammar (line 1, Table 3). Phonological memory skill for Spanish-like stimuli was significantly related to Spanish vocabulary and Spanish grammar, but not to English vocabulary or grammar (line 2, Table 3). English vocabulary was related to English grammar (line 3, Table 3), and Spanish vocabulary was related to Spanish grammar (line 4, Table 3). There was a significant negative cross-language correlation between English vocabulary and Spanish grammar (line 3, Table 3). The cross-language correlation between Spanish vocabulary and English grammar was nonsignificant (line 4, Table 3).

All the zero-order correlations among measures of phonological memory and oral language development are influenced by the common effects of language exposure on those variables. Because exposure to English and Spanish were measured as proportions of input, these two measures are necessarily negatively related. Thus, to the degree that any of the other measures depend on input, the trade-off between the relative amount of exposure to English vs. Spanish pushes within-language correlations in a positive direction and across-language correlations in a negative direction. Marchman et al. (2004) have argued that appropriate estimation of within- and across-language effects requires holding constant the effects of language exposure. In contrast to the pattern of language-specific effects seen in the zero-order correlations, the partial correlations observed in this study suggested both within-language and across-language relations between phonological memory skill and language development.

The problem with these partial correlations, however, is that removing effects of language exposure in assessing relations between phonological memory and language removes exactly the language-specific variance in phonological memory skill that would be the basis of a language-specific effect of phonological memory on language development. We argue that a better test of the hypothesis that language exposure supports the development of language-specific phonological memory skills which, in turn, support language development can be accomplished using hierarchical regression to first remove effects of phonological memory skill in the other language in order to estimate, in the second step, the variance accounted for by language-specific phonological memory skill. A third step provides estimates of the effects of input that are not mediated by language-specific phonological memory skill. The results of those regressions with English vocabulary and grammar and Spanish vocabulary and grammar as outcomes are presented in Tables ​4 and ​5. They suggest that after the shared variance with phonological memory skill for Spanish-like stimuli is removed, variance in phonological memory skill for English-like stimuli uniquely accounts for a significant 35% of the variance in English vocabulary and 27% of the variance in English grammar. Language exposure accounts for an additional significant 25 and 15 percent of the variance in these outcomes, respectively. (The percent of language exposure that is English is a positive predictor of English outcomes and a negative predictor of Spanish outcomes.) After the shared variance with phonological memory skill for English-like stimuli is removed, variance in phonological memory skill for Spanish-like stimuli uniquely accounts for a significant 12 % of the variance in Spanish vocabulary; the relation to Spanish grammar was not significant. Language exposure accounts for an additional 31 and 21 percent of the variance in these outcomes.

Table 4

Hierarchical Regression Analyses Predicting English Vocabulary and Grammar at 25Months (n = 41)

Table 5

Hierarchical Regression Analyses Predicting Spanish Vocabulary and Grammar at 25Months (n = 41)

Discussion

In this study, the relations among Spanish-English bilingually-developing children’s relative amount of exposure to each of their languages, their phonological memory skills in each language (these measured at 22 months), and their productive vocabulary size and grammar in each language (measured at 25 months), were examined in a sample of children who had been exposed to both languages from birth. The study made use of variability in these children’s relative exposure to their two languages to estimate the degree to which phonological memory skills depend on language experience and the degree to which the value of phonological memory to language acquisition might thus also be language-specific.

There was evidence in the data that phonological memory skill is in part dependent on language experience. Relative exposure to English accounted for a significant 25% of the variance in children’s phonological memory for English-like stimuli that was not shared with variance in phonological memory for Spanish-like stimuli; relative exposure to Spanish accounted for 20% of the unique variance in phonological memory for Spanish-like stimuli. Because the stimuli were not real words but were sound sequences that conformed to the phonological properties of each language, we hypothesize that these relations reflect the effect of language exposure resulted in the children’s building of mental representations of the phonemes and phonotactics of each language. These representations are the knowledge-based component of phonological memory skill. This finding and interpretation are consistent with evidence in the literature that children show more accurate nonword repetition for stimuli that are more word-like (Gathercole et al., 1991) and that adults and children (Thorn & Gathercole, 1999) show more accurate nonword repetition for their own language than for a foreign language (Service & Kohonen, 1995).

The strong positive correlation between phonological memory skill for English-like and Spanish-like stimuli (shared variance equaled 64%) suggests that the ability to repeat sequences of phonemes in these two languages is also dependent on a common underlying ability. This common underlying ability likely includes a general auditory memory capacity that does not make use of any particular knowledge base. The correlation between phonological memory for English-like and Spanish-like stimuli may also reflect the fact that there is substantial overlap between the phonemes of English and Spanish, thus the knowledge base that supports phonological memory for English- and Spanish-like stimuli is in part a single knowledge base. The present data do not provide a way to partition the variance that is shared between English and Spanish phonological memory skills into that which depends on a general capacity and that which reflects phonological overlap. Research on bilingual children acquiring languages that differ more in their phonological properties may be able to address this question.

Despite the overlap between English and Spanish phonology, there was evidence of language-specific relations between these children’s phonological memory skills and their vocabulary and grammatical development in each language. That is, 35% of the variance in English vocabulary and 27% of the variance in English grammar were attributable to variance in phonological memory for English-like stimuli, after the variance shared with memory for Spanish-like stimuli was removed; 12% of the variance in Spanish vocabulary was attributable to variance in phonological memory for Spanish-like stimuli, after the variance shared with memory for English-like stimuli was removed. Phonological memory for Spanish-like stimuli did not uniquely account for variance in Spanish grammar, but floor effects on the grammar measure limited its value. Language exposure made additional direct contributions to explaining variance in all outcomes. These findings are consistent with the hypothesis proposed in Hoff et al. (2008), that in addition to direct influences of language exposure on language development, exposure has indirect influences mediated by phonological memory skill.

Although the relation of language exposure to phonological memory and the relation of phonological memory to vocabulary and grammar were the foci of this investigation, the present study also provided data on the relation of language exposure to vocabulary and grammar and data on the within- and across-language relations between vocabulary and grammar. The proportion of children’s home language exposure that was English was a significant positive predictor of their English vocabulary and grammar and a significant negative predictor of their Spanish vocabulary and grammar. These findings are consistent with the results of previous studies of bilingual children (Pearson et al., 1997), with other findings from the larger study from which this sample was drawn (Hoff et al., under review), and with a large body of evidence from the study of monolingual children that the amount of language input children receive predicts the rate of their language development (Hoff, 2006).

The present data also showed strong correlations between vocabulary size and grammar within languages, but not across languages, replicating findings by Marchman et al. (2004) and Conboy and Thal (2006). The process underlying these correlations is a topic of some controversy, but is likely to reflect both mutual influences between vocabulary and grammatical development and the influence of a common cognitive capacity on both (Conbody & Thal, 2006; Dionne, Dale, Boivin, & Plomin, 2003; Marchman et al., 2004; Naigles, Hoff, & Vear, 2009). Last, the present data showed that with effects of language exposure held constant, children’s vocabulary size in English and Spanish were significantly correlated and the grammatical complexity of their English and Spanish were unrelated. The positive relation between English and Spanish vocabulary suggests a common underlying capacity that serves vocabulary development in both languages, and the other findings from this study suggest that phonological memory is likely to be at least a component of that common underlying capacity. The implication of the nonsignificant correlation between English and Spanish grammar is less clear, but, particularly for Spanish, many children were at zero on this measure and that lack of variance may be the explanation for the null finding.

The present findings leave several questions unaddressed. The data are only correlational, and it may be that the influence between phonological and lexical development also operates in the direction of vocabulary growth prompting phonological reorganization as some theories posit (Beckman & Edwards, 2000; Metsala & Walley, 1998). If so, then the relations among language exposure, the development of phonological memory skill, and the growth of vocabulary and grammar are spiraling mutual influences rather than unidirectional, but the language-specificity of those relations still holds.

In sum, the present findings replicate in young bilingual first language learners the relation between phonological memory skill and language development that has been well demonstrated in monolingual development (Gathercole, 2006). This finding, in combination with the finding that the children’s phonological memory for English-like stimuli was related to their relative amount of English exposure, supports the hypothesis suggested by Hoff et al. (2008), that children’s phonological memory skills develop in part as a result of their building phonological representations that support storage of newly-encountered sound sequences that conform to the learned phonological system. The present findings that phonological memory for English-like stimuli uniquely accounted for variance in English language skill and phonological memory for Spanish-like stimuli uniquely accounted for variance in Spanish language skill supports the proposed function of phonological memory as temporary storage of individual exemplars of words and sentences as they are encountered, which in turn provides a database for the abstraction of lexical items and grammatical patterns (Gathercole, 2006; Speidel, 1993). Last, the findings that language exposure showed language-specific relations to phonological memory and to language development and that phonological memory partially mediated the effect of exposure on development suggests that language exposure benefits language development both directly, as the source of the database on which learning depends, and indirectly as an influence on the capacity to temporarily store that database.

Acknowledgments

This research was supported by grants from the National Institute of Child Health and Human Development (HD054427 to Erika Hoff). This work is based on a thesis completed in partial fulfillment of the requirement for a Master’s degree in Psychology at Florida Atlantic University.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Marisol Parra, Department of Psychology, Florida Atlantic University.

Erika Hoff, Department of Psychology, Florida Atlantic University.

Cynthia Core, Department of Communication Sciences and Disorders, Florida Atlantic University.

References

• Adams AM, Gathercole SE. Phonological working memory and speech production in preschool children. Journal of Speech and Hearing Research. 1995;38:403–414. [PubMed] [Google Scholar]
• Adams AM, Gathercole SE. Phonological working memory and spoken language development in young children. Quarterly Journal of Experimental Psychology. 1996;49A:216–233. [Google Scholar]
• Adams AM, Gathercole SE. Limitations in working memory: Implications for language development. International Journal of Language and Communication Disorders. 2000;35:95–116. [PubMed] [Google Scholar]
• Baddeley AD, Gathercole SE, Papagno C. The phonological loop as a language learning device. Psychological Review. 1998;105:158–173. [PubMed] [Google Scholar]
• Beckman ME, Edwards J. The ontogeny of phonological categories and the primacy of lexical learning in linguistic development. Child Development. 2000;71:240–249. [PubMed] [Google Scholar]
• Bialystok E. Metalinguistic aspects of bilingual processing. Annual Review of Applied Linguistics. 2001;21:169–181. [Google Scholar]
• Bialystok E, Feng X. Language proficiency and its implications for monolingual and bilingual children. In: Durgunoğlu A, editor. Challenges in language and literacy issues. Guildford Press; (in press) [Google Scholar]
• Bosch L, Sebastian-Gallés N. Simultaneous bilingualism and the perception of a language specific vowel contrast in the first year of life. Language and Speech. 2003;46:217–244. [PubMed] [Google Scholar]
• Burns TC, Yoshida KA, Hill K, Werker JF. The development of phonetic representation in bilingual and monolingual infants. Applied Psycholinguistics. 2007;28:455–474. [Google Scholar]
• Cheung H. Nonword span as a unique predictor of second-language vocabulary learning. Developmental Psychology. 1996;32:867–873. [Google Scholar]
• Chiat S, Roy P. The preschool repetition test: An evaluation of performance in typically developing and clinically referred children. Journal of Speech Language and Hearing Research. 2007;50:429–443. [PubMed] [Google Scholar]
• Coady JA, Aslin RN. Young children’s sensitivity to probabilistic phonotactics in the developing lexicon. Journal of Experimental Child Psychology. 2004;89:183–213. [PMC free article] [PubMed] [Google Scholar]
• Coady JA, Evans JL. Uses and interpretations of non-word repetition tasks in children with and without specific language impairment (SLI) International Journal of Language and Communication Disorders. 2008;43:1–40. [PMC free article] [PubMed] [Google Scholar]
• Conboy BT, Thal DJ. Ties between the lexicon and grammar: Cross sectional and longitudinal studies of bilingual toddlers. Child Development. 2006;77:712–735. [PubMed] [Google Scholar]
• Dionne G, Dale PS, Boivin M, Plomin R. Genetic evidence for bidirectional effects of early lexical and grammatical development. Child Development. 2003;74:394–412. [PubMed] [Google Scholar]
• Edwards J, Beckman ME, Munson B. The interaction between vocabulary size and phonotactic probability effects on children’s production accuracy and fluency in nonword repetition. Journal of Speech, Language, and Hearing Research. 2004;47:421–436. [PubMed] [Google Scholar]
• Fenson L, Dale PS, Reznick JS, Thal D, Bates E, Hartung JP, et al. The MacArthur Communicative Development Inventories: User’s guide and technical manual. Singular Publishing Group; San Diego, CA: 1993. [Google Scholar]
• French LM, O’Brien I. Phonological memory and children’s second language grammar learning. Applied Psycholinguistics. 2008;29:463–487. [Google Scholar]
• Gathercole SE. Nonword repetition and word learning: The nature of the relationship. Applied Psycholinguistics. 2006;27:513–543. [Google Scholar]
• Gathercole SE, Adams AM. Phonological working memory in very young children. Developmental Psychology. 1993;29:770–778. [Google Scholar]
• Gathercole SE, Hitch GJ, Service E, Martin AJ. Phonological short term memory and new word learning in children. Developmental Psychology. 1997;33:966–977. [PubMed] [Google Scholar]
• Gathercole SE, Willis CS, Emslie H, Baddeley AD. The influence of number of syllables and wordlikeness on children’s repetition of nonwords. Applied Psycholinguistics. 1991;12:349–367. [Google Scholar]
• Gathercole SE, Willis CS, Emslie H, Baddeley AD. Phonological memory and vocabulary development during the early school years: A longitudinal study. Developmental Psychology. 1992;28:887–898. [Google Scholar]
• Goldstein BA. Phonological development and disorders. In: Goldstein BA, editor. Bilingual language development and disorders in Spanish-English speakers. Paul H. Brookes Publishing Co.; Baltimore: 2004. pp. 259–286. [Google Scholar]
• Hoff E. How social contexts support and shape language development. Developmental Review. 2006;26:55–88. [Google Scholar]
• Hoff E, Core C, Bridges K. Nonword repetition assesses phonological memory and is related to vocabulary development in 20- to 24- month-olds. Journal of Child Language. 2008;35:1–14. [PubMed] [Google Scholar]
• Hoff E, Core C, Place S, Rumiche R, Señor M, Parra M. Dual Language Exposure and Early Bilingual Development. 2010 Manuscript submitted for publication. [PMC free article] [PubMed] [Google Scholar]
• Holm A, Dodd B. A longitudinal study of the phonological development of two Cantonese-English bilingual children. Applied Psycholinguistics. 1999;20:349–376. [Google Scholar]
• Jackson-Maldonado D, Thal DJ, Fenson L, Marchman V, Newton T, Conboy B. El Inventario del Desarrollo de las Habilidades Comunicativas: User’s guide and technical manual. Paul H. Brookes Publishing Co.; Baltimore, MD: 2003. [Google Scholar]
• MacDonald MC, Christiansen MH. Reassessing working memory: Comment on Just and Carpenter (1992) and Waters and Caplan (1996) Psychological Review. 2002;109:35–54. [PubMed] [Google Scholar]
• Marchman VA, Martínez-Sussmann C, Dale PS. The language-specific nature of grammatical development: Evidence from bilingual language learners. Developmental Science. 2004;7:212–224. [PubMed] [Google Scholar]
• Messer MH, Leseman PPM, Boom J, Mayo AY. Phonotactic probability effect in nonword recall and its relationship with vocabulary in monolingual and bilingual preschoolers. Journal of Experimental Child Psychology. 2010;105:306–323. [PubMed] [Google Scholar]
• Metsala JL, Walley AC. Spoken vocabulary growth and the segmental restructuring of lexical representations: Precursors to phonemic awareness and early reading ability. In: Metsala JL, Ehri LC, editors. Word recognition in beginning literacy. Erlbaum; Mahwah, NJ: 1998. pp. 89–102. [Google Scholar]
• Munson B, Kurtz BA, Windsor J. The influence of vocabulary size, phonotactic probability, and wordlikeness on nonword repetitions of children with and without specific language impairment. Journal of Speech, Language, and Hearing Research. 2005;48:1033–1047. [PubMed] [Google Scholar]
• Naigles LR, Hoff E, Vear D. Flexibility in early verb use: Evidence from a multiple-n diary study. Monographs of the Society for the Research in Child Development. 2009;74(2, Serial No. 293) [PubMed] [Google Scholar]
• Oller DK, Cobo-Lewis AB, Eilers RE. Phonological translation in bilingual and monolingual children. Applied Psycholinguistics. 1998;19:259–278. [Google Scholar]
• Paradis J. Do bilingual two-year-olds have separate phonological systems? International Journal of Bilingualism. 2001;5:19–38. [Google Scholar]
• Pearson BZ, Fernández VL, Lewedeg V, Oller K. The relation of input factors to lexical learning by bilingual children. Applied Psycholinguistics. 1997;18:41–58. [Google Scholar]
• Place S, Hoff E. Language use in bilingual homes. Poster presented to the Conference on Human Development; New York. April 9 – 11.2010. [Google Scholar]
• Schneider W, Bjorklund DF, Maier-Brückner W. The effects of expertise and IQ on children’s memory: When knowledge is, and when it is not enough. International Journal of Behavioral Development. 1996;19:773–796. [Google Scholar]
• Service E, Konohen V. Is the relation between phonological memory and foreign language learning accounted for by vocabulary acquisition? Applied Psycholinguistics. 1995;16:155–172. [Google Scholar]
• Shriberg LD, Lohmeier HL, Campbell TF, Dollaghan CA, Green JR, Moore CA. A nonword repetition task for speakers with misarticulations: The syllable repetition task (SRT) Journal of Speech Language and Hearing Research. 2009;52:1189–1212. [PMC free article] [PubMed] [Google Scholar]
• Snowling MJ. Phonemic deficits in developmental dyslexia. Psychological Research. 1981;43:219–234. [PubMed] [Google Scholar]
• Speidel GA. Phonological short-term memory and individual differences in learning to speak: A bilingual case study. First Language. 1993;13:69–91. [Google Scholar]
• Squires J, Potter L, Bricker D. Ages and Stages Questionnaire: A Parent-Completed Child Monitoring System. 2nd Edition Brooks Publishing; Baltimore: 1999. [Google Scholar]
• Thorn SC, Gathercole SE. Language-specific knowledge and short-term memory in bilingual and non-bilingual children. Quarterly Journal of Experimental Psychology: Human Experimental Psychology. 1999;52:303–324. [PubMed] [Google Scholar]
• Vallar G. Memory systems: The case of phonological short-term memory. A festschrift for Cognitive Neuropsychology. Cognitive Neuropsychology. 2006;23:135–155. [PubMed] [Google Scholar]